A strong application helps you see what you had not yet seen in yourself, and where to go next.

We put her experiences back on the same table: childhood separation, girls' basketball, photography across many countries, an economics club, an education nonprofit, business internships, and a cross-border donation effort during the pandemic. These were parallel parts of her life; none had been undertaken to manufacture an education-policy application.

What did repeat was her concern with access to education and how people grow. A small student club developed into an organisation spanning multiple schools. As it expanded, funding and teacher shortages exposed the limits of individual execution and led her to ask what institutional change would require.

The four schools did not receive one recycled essay. We established a shared core, then rebuilt the material around what each programme needed to evaluate:

• For Harvard, the story moved from personal experience to educational practice, then to the limits of that practice and the need for policy training. Her business background appeared as the judgement she used to solve resource, operational, and execution problems.
• For Chicago, academic potential mattered more. Basketball, photography, and nonprofit work became evidence of how she observes people, understands institutions, and develops research questions from lived reality.

She was admitted to all four programmes and chose Harvard. The Harvard name helped her secure more consulting interviews; her business foundation and case-interview performance converted those opportunities. She ultimately joined an MBB firm.

We revisited each internship and asked what she had noticed, doubted, and used as evidence. During one private-equity diligence project, a food manufacturer claimed to have automated production. She kept pressing for costs, R&D evidence, and actual deployment, but received no verifiable answer. She ultimately recommended against the investment. That detail transformed "participated in diligence" into visible professional judgement.

The four internships represented four positions relative to numbers and the business itself: understanding accounts from inside a company, reading financial statements and valuations from outside, testing whether management's story was credible, and finally using data directly inside an operating business.

For accounting, we emphasised her scrutiny of whether numbers could be trusted. For business analytics, we showed how she entered the operating context and found actionable decisions in the data. The same experiences were not forced into two artificial personas; they revealed two real dimensions of the same candidate.

Accounting offers from Cornell, Michigan, and UIUC, with a scholarship from UIUC; business analytics offers from Johns Hopkins Carey and Toronto Rotman.

The first step was not drafting an essay. It was going back to understand the internship properly. We asked her to speak with the hotel manager again. The everyday operating facts included two hotels of the same brand competing near one transport hub; weekday occupancy reaching 95% and weekend occupancy falling to 60%; holiday traffic at 1.5 times the normal level; different weekday and weekend pricing; a customer mix split among members, third-party platforms, and corporate accounts; and 10.7% year-on-year occupancy growth.

She believed she had done nothing because she saw only routine operations, not the analytical questions contained within them.

The same facts became a real analytical setting: two comparable hotels; dynamic pricing in response to demand; and customer segmentation across weekday and weekend patterns.

Once the operating context was verified, our subject-matter mentor helped her identify researchable questions. She then completed the underlying work: customer segmentation, dynamic pricing, time-series demand forecasting, channel analysis, and dashboards. The business situation was real; the mentor's role was to help her select appropriate methods and represent only the analysis she actually completed.

Two summers at the same hotel were organised as a progression. The first focused on descriptive analysis, seasonal patterns, and KPI dashboards. After more coursework, the second moved into modelling. The application could therefore show growth from description to prediction and from analysis to implementation.

The strongest evidence was not a model result but her process of changing methods: linear regression exposed nonlinearity; ARIMA captured ordinary trends but missed event-driven spikes; event calendars and weather were then introduced as exogenous variables. The revisions showed how she recognises a model's limits and improves it.

Admitted to the University of Chicago's Applied Data Science programme.